Using Machine Learning to Predict Serial Killers’ Vital Goal
Bolun Ha · 2021
Serial killers' criminal behavior and goal are sometimes challenging to find from criminals or very few cases before they are arrested. It is difficult to analyze some unique relationships through massive data that is a challenge to disclose the facts to the public. According to the study from Bartol, C. R., & Bartol, A. M. (2017)[1], there are still more than 30% of criminals who clearly commit crimes in a wider geographical area. Serial killers are mainly men, and they tend to prefer one sex over another. Therefore, through data visualization and hypothesis testing, we can get a strong linear relationship between killers and victims. This paper analyzes a data set based on a sample of American serial killers from the Radford/FGCU Serial Killer Database, visualizes American serial killers' behaviors and uses machine learning to classify and predict whether the serial killers have a specific goal of victims.